This graph says it all: the 2016 recession was already here

In August 2016 the Bureau of Economic Analysis put American GDP growth at 1.2 per cent, and this page made a claim that no official body would make for another two years, if ever: that the United States was already in recession. The evidence was a single chart, drawn by 720Global, of seventy years of growth with recessions shaded, and a red line at 1.2 per cent. At almost every point in those seventy years where growth touched that line, the economy was in a shaded area. This page keeps the chart’s argument and the two things it showed beyond the immediate call.

The line

The first thing the chart showed was that 1.2 per cent is not a growth rate on the way to a recession; historically it is a rate inside one. The two exceptions since 2009 proved the rule. On both occasions when growth touched the line and no recession followed, the Federal Reserve had responded with a large round of quantitative easing, and growth had staggered back up on the new money. The third round, dosed monthly rather than in a single hit, produced the stepped rise visible on the chart from 2012 to 2014. When it ended in October 2014, the line resumed its decline toward the red line at a steady angle, exactly as this site had said in early 2014 that it would once the support was withdrawn.

The longer trend

The second thing the chart showed was larger than 2016. The economy came out of the war growing at 12 per cent. For the next thirty years each expansion peaked at around 8. For the twenty years after that, about 4. In the recovery after 2009, with the largest monetary stimulus in history and interest rates at zero, it could not reach 3. Part of this is arithmetic: a larger economy is harder to grow by a given percentage, as it is easier to double one’s speed at ten miles an hour than at fifty. But the size of the decline, and the size of the stimulus required to achieve less each time, point to something heavier.

The candidates can be checked against the chart. Taxes are not it: the fastest growth came in decades when income and capital-gains rates were far higher than in 2016, and thirty years of cutting them had coincided with the slowest. Free trade is not it either: the steepest declines in trend growth came before the trade agreements of the 1990s, though it is fair to say the agreements did not reverse them. What is consistent across the whole chart is debt. Every recession was fought with borrowing and none of the borrowing was repaid in the expansions that followed, so each expansion started with more ballast than the last. An economy that has grown by not paying for its growth eventually finds that the interest on the past consumes the present.

Why nobody had declared it

Recessions are declared after the fact. The two-quarter shorthand cannot be applied until the second quarter’s numbers are in, which is seven months at the earliest, and the official dating committee usually takes a year or more. The economy is often out of a recession before anyone is told it was in one. That is why the claim made here was not that a recession had been announced but that the conditions of one were present, and that the announcement, if it came, would come late.

The forecast that followed was in three parts: that the Fed would eventually resort to one more large round of stimulus; that it would come too late, because the Fed needed its recovery to have worked and would resist admitting otherwise; and that whatever it bought would be less than the last round bought, because the returns were diminishing on the chart for everyone to see.

Later note. The NBER did not date a recession in 2016. Growth recovered to 2.9 per cent in 2018 on a large tax cut and deficit, and the next recession, in 2020, was caused by a pandemic and answered with the largest stimulus yet, which is the third part of the forecast in a form nobody predicted. The trend line on the chart, extended through 2024, had not turned up.